A complete sample of LSP blazars fully described in γ -rays. New γ -ray detections and associations with Fermi-LAT
AAstronomy & Astrophysics manuscript no. Complete-Sample-LSP-blazars-fully-described-in-gamma-rays c (cid:13)
ESO 2018April 12, 2018
A complete sample of LSP blazars fully described in γ -rays New γ -ray detections and associations with Fermi-LAT B. Arsioli , , and G. Polenta Science Data Center della Agencia Spaziale Italiana, SSDC - ASI, Rome, Italy Instituto de F´ısica Gleb Wataghin, UNICAMP, R. S´ergio Buarque de Holanda 777, 13083-859 Campinas, Brazil ICRANet-Rio, CBPF, Rua Dr. Xavier Sigaud 150, 22290-180 Rio de Janeiro, Brazil ASI - Agenzia Spaziale Italiana, via del Politecnico snc, 00133 Roma, Italye-mail: [email protected],[email protected] e-mail: [email protected]
Preprint online version: April 12, 2018
ABSTRACT
Context.
We study the γ -ray and broadband spectral energy distribution (SED) properties of a complete sample of 104bright, radio-selected low-synchrotron peaked (LSP) blazars, which have well-characterized SEDs from radio to X-rays.Most of the sources have already been detected in the γ -ray band by Fermi-LAT, however almost 20% of these blazarshave no counterpart in any of the Fermi catalogs published so far. Aims.
Using the Fermi Science Tools, we look for γ -ray emission for those objects not yet reported in any Fermi-LATcatalog, finding new detections and associations. We then study the multifrequency SED for all sources in our sample,fitting their synchrotron (Syn) and inverse Compton (IC) components. A complete sample of LSP blazars with a fulldescription in γ -ray is unique. We use this sample to derive the distribution of the Compton dominance (CD) alongwith population properties such as Syn and IC peak power, and frequency distributions. Methods.
We performed a binned likelihood analysis in the 0.3-500 GeV energy band with Fermi-LAT Pass 8 data,integrating over 7.5 years of observations. We studied γ -ray light curves and test statistic (TS) maps to validate newdetections and associations, thereby building a better picture of the high-energy activity in radio-selected LSP blazars.We fit the IC component for the new detections using all data at our disposal from X-rays to GeV γ -rays, enhancingthe amount of information available to study the Syn to IC peak-power correlations. Results.
We deliver a unique characterization in γ -rays for a complete sample of LSP blazars. We show that threepreviously unidentified 3FGL sources can be associated with blazars when using improved γ -ray positions obtainedfrom TS maps. Six previously unreported γ -ray sources are detected at TS >
20 level, while another three show TSvalues between 10-20. We evaluate two cases in which source confusion is likely present. In four cases there is nosignificant γ -ray signature when integrating over 7.5 years. Short-lived flares at ∼ ≈ Key words.
Gamma rays: Galaxies – Galaxies: Active – Radiation mechanism: Non-thermal
1. Introduction
Blazars are a particular class of jetted active galactic nu-clei (AGN), corresponding to the very few cases where jetsare pointing close to our line of sight (Padovani et al. 2017).These objects are known for having a rather unique spectralenergy distribution (SED) often characterized by the pres-ence of two nonthermal bumps in the log( ν f ν ) versus log( ν )plane, which extends along the whole electromagnetic win-dow from radio up to TeV γ -rays. Also known for theirrapid and large amplitude spectral variability, usually theobserved radiation shows extreme properties owing to therelativistic nature of the jets, which result in amplificationeffects. Blazars are relatively rare; there are ∼ ν f ν ) vs. log( ν ) plane is asso-ciated with the emission of Syn radiation owing to rela-tivistic electrons moving through the magnetic field of thecollimated jet. The second peak is usually understood asa result of inverse Compton (IC) scattering of low energyphotons to the highest energies, by the same relativisticelectron population that generates these Syn photons (syn-chrotron self Compton model; SSC). The seed photons un-dergoing IC scattering can also come from outside regions,such as the accretion disk and broad line region, and canadd an extra ingredient (external Compton models; EC) formodeling the observed SED.Since the peak-power associated with the Syn bump tellus at which frequency ( ν Synpeak ) most of the AGN electro-magnetic power is being released, the parameter log( ν Synpeak )has been extensively used to classify blazars. FollowingPadovani & Giommi (1995); Abdo et al. (2010), objects a r X i v : . [ a s t r o - ph . H E ] A p r . Arsioli and G. Polenta: A complete sample of LSP blazars fully described in γ -rays with log( ν Synpeak ) < > γ -rays. Nevertheless a relatively large percentage of LSPs stilllack detection by Fermi-LAT. The fact that some of theseLSPs are relatively bright in radio and show hints of dis-tinct optical polarization properties (Angelakis et al. 2016)has motivated arguments about the existence of a specificclass to represent the so-called γ -ray quiet blazars; the lat-ter have a relatively lower polarization fraction. Therefore,current evidence gives us a hint of the jet condition re-garding γ -ray undetected blazars, showing they might beconnected to relatively less magnetized jets. In particular,Blinov et al. (2015) probed optical polarization swings inconnection to periods of enhanced activity in γ -rays, there-fore in connection with magnetic field strength and orderingwithin the jet structure. But the observations of orphan γ -ray flares, which have no counterparts at the low-energysite (optical to X-ray), are a standing challenge for cur-rent SSC and EC models when trying to describe blazarvariability (Potter 2018). Alternative models such as theRing of Fire (MacDonald et al. 2017) discuss evidence fora sheath of plasma surrounding the spine of the jet, pro-ducing a dominant IR photon field that would undergo ICscattering. Despite that, other works recognize γ -ray unde-tected blazars as MeV peaked (Paliya et al. 2017), whichcould be out of reach for Fermi-LAT.As known, all Fermi-LAT official catalogs are blindwith respect to other wavebands, meaning that informa-tion about sources detected in other energy bands are nottaken into account. Each new γ -ray detection must meet theconservative requirement of being observed at least as a 5 σ excess compared to the expected background. This is usu-ally incorporated in the so-called test statistic (TS) param-eter (more on sec. 3) translating to a TS >
25 requirementfor acceptance of new γ -ray sources. This was necessary toavoid spurious detections and misleading associations, es-pecially pre-Fermi-LAT, when the main population of γ -rayemitters was still to be identified.The situation now has changed since the astrophysicscommunity already recognize blazars as the main popu-lation of extragalactic γ -ray emitters. Searches for γ -rayemission in samples of previously known blazars can relaxthe TS >
25 requirement, as discussed in Arsioli & Chang(2017). These authors tested the existence of γ -ray signa-tures by including more information compared to a pureblind approach and they indeed detected 150 new γ -raysources. The detection of new sources, together with resolv-ing cases of source confusion, should not be neglected givenits direct impact on statistical γ -ray properties of specificpopulations, particularly for small samples.Especially in case of LSP blazars, the percentage of γ -ray undetected sources is higher. On average, those blazarshave steep γ -ray photon spectral index (Ackermann et al.2015b) that can compromise the detection of high redshiftsources since the γ -ray SED shifts to lower frequencies inthe observer rest frame. Also, the absorption of very highenergy (VHE) γ -ray owing to the interaction with the ex-tragalactic background light (EBL; Franceschini, A. et al.2008) may hinder observations with Fermi-LAT. In addi-tion to those possible complications, intrinsic jet proper- ties (such as the Doppler and beaming factor δ ) and thedominant IC regime (either SSC or EC) may also have alarge influence on the Fermi-LAT detectability of radio-loudblazars. In fact, Lister et al. (2009) showed that the γ -raysources detected during the first three months of Fermi-LAT operations are associated with the largest apparentjet speeds (therefore the largest bulk Lorentz factors) asdeduced from radio measurements with the Very LargeBaseline Array (VLBA). Also, Lister et al. (2015) showedthat the γ -ray detection of LSP blazars is relatively lesslikely when the Lorentz factor is low and the Syn peak isbelow 10 . Hz. In fact, Sec. 4.5 presents three examples ofradio-loud blazars with relatively bright ν f ν Syn peaks thatare detectable only during short flare episodes. Those casescould provide hints of the mechanisms behind γ -ray flaresand are interesting targets for upcoming missions such asthe MeV dedicated e-Astrogram (Tatischeff et al. 2016; DeAngelis et al. 2017).Motivated by the possibility of unveiling new γ -raysources, we use a complete sample of radio-loud blazars,and consider all cases not yet detected at GeV band for alikelihood analysis with the Fermi Science Tools. Our ap-proach shows that most of the previously γ -ray undetectedsources are actually detectable when integrating over 7.5years of observations or during short periods of their flar-ing states. We present γ -ray light curves adopting a onemonth time bin, showing that currently undetected blazarsmay introduce a dynamic high-energy component to the skywith numerous short periods of γ -ray activity. This mightbuild a considerable portion of the extragalactic γ -ray back-ground (EGB; Ackermann et al. 2015a, 2013) whose originis still being debated (Fornasa & S´anchez-Conde 2015; DiMauro & Donato 2015; Di Mauro et al. 2014; Ackermannet al. 2016a). In addition, undetected blazars may addanisotropic contributions to the EGB content especially atsmall angular scales(Ackermann et al. 2012; Cuoco et al.2012; Inoue 2014) thus potentially affecting the search fordark matter annihilation or decay signatures in connec-tion to diffuse γ -ray emission from extragalactic large-scalestructures (Ando et al. 2014; Prokhorov & Churazov 2014).Indeed, our new detections may complement current un-derstanding of the EGB origin especially for the MeV bandwhere LSP and ISP blazars are more relevant.Our detailed search for new γ -ray sources also providesa better description of the IC component for many blazarsthat so far had limited or no γ -ray data available, thusresulting in a measurement of Syn and IC peak-power pa-rameters for nearly all sources (except for five nondetec-tions). Since this radio sample includes the brightest andwell-characterized LSP blazars, we study population prop-erties such as the distribution of Syn and IC peak frequen-cies and peak power, and the distribution of log( ν IC /ν Syn )and log(Compton dominance).
2. Sample description
We consider the complete sample of radio-loud AGN thatwas studied in detail by Planck Collaboration et al. (2011),consisting of 104 northern and equatorial sources with dec-lination larger then -10 ◦ and flux density at 37 GHz exceed-ing 1 Jy as measured with the Mets¨ahovi radio telescope.We refer to those sources as the Radio-Planck sample andlist these sources in table 5 showing their blazar names
2. Arsioli and G. Polenta: A complete sample of LSP blazars fully described in γ -rays Table 1.
List of new γ -ray detections and associations for the Radio-Planck sources, following a case by case descriptionin Sec. 4. Columns Source and Redshift list the blazar name and distance as from Massaro et al. (2015). We notethat Table 5 brings an extra name reference as from NVSS Radio Catalog (Condon et al. 1998) given that each blazarhas a well-defined radio counterpart. Columns log( ν Synpeak ) and log( ν Synpeak ) refer to the peak frequency from Syn and ICcomponents measured in Hz. Columns log( νf Synν ) and log( νf ICν ) correspond to the Syn and IC peak-power measured inerg/cm /s. Column N list the pre-factor as of eq. 1 in units of ph/cm /s/MeV when assuming E = 1000 MeV. ColumnΓ list the photon spectral index as of eq. 1. Column TS list the test statistic values associated with the γ -ray signaturefor each source when integrating over 7.5 years of Fermi-LAT PASS8 observations. All model parameters were derivedusing the Fermi Science Tools assuming a power-law spectrum within the 0.3-500 GeV energy band. New detections atTS >
20 level are denoted with an “a” flag; new association from better positioning of 3FGL sources are given a “b”flag, faint detections with TS between 10-20 are given a “c” flag; sources only detectable during short flare episodes aregiven a “d” flag (for those cases, TS values refer to the bright state integrated over one month); and sources that couldbe confused with a second γ -ray emitter are given an “e” flag. Source Redshift log( ν Synpeak ) log( ν f Synν ) log( ν ICpeak ) log( ν f ICν ) N (10 − ) Γ TS Pass . yrs a ± ± a ± ± a ± ± a ± ± a ? -11.7 3.8 ± ± a ± ± b ± ± b ± ± b ± ± c ? -11.7 2.7 ± ± c ± ± c ? -12.1 1.6 ± ± d ± ± ∗ d ? -10.3 20.9 ? -10.6 6.3 ± ± ∗ d ± ± ∗ e ? -12.3 9.1 ± ± from 5BZcat (Massaro et al. 2015), and their NVSS radiocounterpart. Out of these sources, 83 have a confirmed γ -ray counterpart in at least one of the Fermi-LAT (Atwoodet al. 2009) catalogs, i.e., 1FGL, 2FGL, and 3FGL (Abdoet al. 2010; Ackermann et al. 2011; Acero et al. 2015) or2FHL and 3FHL (Ackermann et al. 2016b; Ajello et al.2017). However, the nondetection by Fermi-LAT of theremaining 21 equally bright radio-loud AGN with simi-lar radio-to-optical SED is intriguing, rising discussions onboth the nature of the high-energy emission in blazars andthe efficiency of Fermi-LAT in solving faint γ -ray sources(Lister et al. 2015). It has been argued that LSP blazarswith ν peak < . Hz may show a typical IC peak below0.1 GeV out of the Fermi-LAT sensitivity bandwidth (0.1-500 GeV), so that we can only probe the very end of the ICcomponent (Paliya et al. 2017). γ -ray analysis of undetected sources In a blind analysis, the spectral parameters of a hypothet-ical source, such as normalization and photon spectral in-dex, and also the source position itself (R.A. and Dec.) areall free parameters that have to be optimized during thedata analysis. We might be able to reduce the uncertainty,however, with respect to position since we know multiple γ -ray blazar candidates that have already been identifiedfrom multifrequency observations from radio up to X-rays.Actually, a total of 21 sources in our Radio-Planck sample National Radio Astronomy Observatory Very Large ArraySky Survey, the so-called NVSS Catalog do not have a γ -ray counterpart in previous Fermi FGLcatalogs. Those constitute a set of 21 seed fixed positions,for which we test the existence of relevant γ -ray signatures.This method has been successfully tested for a set of 400 γ -ray candidates (Arsioli & Chang 2017) preselected froma sample of HSP blazars, resulting in 150 new detections.We performed a likelihood analysis integrating over 7.5years of Fermi-LAT observations in the 0.3-500 GeV bandusing Pass 8 data release (Atwood et al. 2013), and assum-ing the γ -ray spectrum of a new source could be describedby a power-law model asdNdE = N (cid:16) EE (cid:17) − Γ , (1)where E is a scale parameter (also known as pivot en-ergy), N is the pre-factor (normalization) correspondingto the flux density in units of ph/cm /s/MeV at the pivotenergy E , and Γ is the photon spectral index for the en-ergy range considered. Both Γ and N are set as free pa-rameters and further adjusted by the fitting routine gtlike .Source positions and E = 1000 MeV are set as fixed pa-rameters, therefore constants for the likelihood analysis.In the source-input xml file, all sources within 10 ◦ fromthe candidate had both Γ and N parameters flagged asfree . Therefore the 3FGL models of these sources, whichare based on four years of observations were adjusted, since In this regard we are following recommendations from FermiScience Tools user guide https://fermi.gsfc.nasa.gov/ssc/data/analysis/scitools/binned_likelihood_tutorial.html , which recommends setting free parameters at least within7 ◦ from the source of interest. This is a consequence of the3. Arsioli and G. Polenta: A complete sample of LSP blazars fully described in γ -rays we integrated over 7.5 years of data. This particular choiceincreases the computational burden of the analysis, but it iscrucial for adapting the model maps to the extra 3.5 yearsof exposure that is being considered. Results are shown inTable 1, listing only the 16 cases that had no known coun-terpart from previous γ -ray catalogs and which we now de-scribe case by case in this work. Also we note that Table 5holds the description for the entire Radio-Planck sample.Since we are dealing with sources that are predomi-nantly LSP blazars, we should expect them to have steep γ -ray SED (in log( ν f ν ) versus log( ν ) plane) as observed fromthe correlation in Γ versus log( ν Synpeak ) plane (Acero et al.2015), where LSP sources dominate the Γ > . represents the flux at 1 GeV (given our choicefor E ). Therefore new detections are expected to be onthe border or below Fermi-LAT four year sensitivity limit ,which is E dN / dE ≈ . × − erg/cm /s at 1 GeV,i.e., dN / dE ≈ . × − ph/cm /s/MeV. Following ourresults as reported in Table 1, few cases have N abovethe the Fermi-LAT four year sensitivity level. In particular5BZQJ 0359+5057 and 5BZUJ 0241-0815 have larger N ,which is clearly a consequence of an enhanced γ -ray activ-ity reported just after 2013 (out of the integration periodused to build the 3FGL catalog) as discussed in Sec. 4.1 and4.5. Another three sources flagged with “b” have N closeto the 3FGL sensitivity border; those are actually detectedin the 3FGL catalog but with large position uncertainty,which lead them to be unassociated in previous catalogs.Positive γ -ray signatures were first evaluated based ontest statistics (TS) values as defined by Mattox et al.(1996): TS = − (cid:16) L (no source) L (source) (cid:17) , where L (no source) is thelikelihood of observing a certain photon count for a modelwithout the candidate source (the null hypothesis), andL (source) is the likelihood value for a model with the addi-tional candidate source at the given location. The reportedTS values correspond to a full band fitting, which con-strains the whole spectral distribution along 0.3-500 GeVto vary smoothly with energy and assuming no spectralbreak. Considering that we have a good description of theGalactic and of the extragalactic diffuse components, thisis a measure of how clearly a source emerges from the back-ground, also assessing the goodness of free parameters fit.A TS ≈
25 is equivalent to a 4-5 σ detection, depend-ing on the strength of the background in the region (Abdoet al. 2010), and only cases with TS >
25 are considered bythe Fermi-LAT team as a positive detection of a point-likesource. Following the discussion on Arsioli & Chang (2017),we analyzed γ -ray signatures down to TS=10, which arespatially consistent with blazars already known from otherenergy bands and double checked these with TS maps.A TS map consists of a pixel grid where the existenceof a point-like source is tested for each pixel, and each gridbin is evaluated using a likelihood analysis . Since the PSF large point spread function (PSF), especially at low-energythreshold, which can overlap with nearby sources. Therefore, inorder to get a confident description of a particular source, weproperly fit and adjust the whole environment around it. For a complete description of the Fermi Science Tools, check: https://fermi.gsfc.nasa.gov/ssc/data/analysis . improves with energy, we worked with E >
500 MeV photonsto help us determine the TS peak position with better pre-cision than working at lower Fermi-LAT bandwidth (downto 100 MeV). Thus the map alone tests the existence of apoint-like source emerging from a flat low-TS background.We enhanced the γ -ray characterization of the Radio-Planck sample by ≈ γ -ray signature,which is probably related to high redshift, IC peak at MeVband, and low galactic latitude hindering the detection.For the light curves, we usually considered a time binof approximately 30 days, and estimated the correspondingflux and errors only when the TS per bin was larger than4.0. When this condition was not satisfied, an upper limit tothe flux was calculated using the integral method (providedby the Fermi Science Tool), which takes into account thebackground level and spectral properties of the test source.
4. New γ -ray detections, validation, andassociation We present new γ -ray signals down to TS ≈
10 level. Wealso describe one case of source confusion, solve three casesof unassociated 3FGL sources, and comment on the nonde-tections as well. Table 1 shows the power-law parametersresulting from the fit in the 0.3-500 GeV energy band, to-gether with redshift of the counterpart, Syn, and IC peakfrequency ( ν peak ), and flux density ( ν f ν ) for all cases stud-ied. For each source, we present a TS map together withthe γ -ray SED with a polynomial fit for both Syn and ICcomponent. Following Giommi et al. (2012b), when fittingthe nonthermal component we also account for optical andUV thermal emission due to accretion using the compositeoptical spectrum built by Vanden Berk et al. (2001). Thisthermal template is based on 2200 optical spectra of radio-quiet quasars (QSOs) taken from the SDSS database andits expected soft X-ray emission, from Grupe et al. (2010).The TS maps are calculated considering only photonswith E >
500 MeV, which is a good choice to evaluatethe TS spatial distribution for these radio sources, sincethey usually have a photon spectral index in the range2.0 < Γ < > γ -ray sig-natures. We call attention to the fact that for most caseswe are dealing with relatively faint γ -ray sources, thereforethe TS distribution could peak at a position slightly offsetfrom it counterpart. An offset is expected given the uncer-tainty introduced by the large PSF, which is on the orderof 1.4 ◦ at 500 GeV (Atwood et al. 2013) and 0.81 ◦ at 1 GeVlevel. Following the discussion in Abdo et al. (2010) whatis important to ensure a proper match is that the 68% con-finement region for the γ -ray signature should enclose thecounterpart blazar. This is explicitly shown case by case tohelp validate our new detections and associations.We show the light curves with monthly bins for thosecases for which we identify a γ -ray flaring activity during
4. Arsioli and G. Polenta: A complete sample of LSP blazars fully described in γ -rays F l u x ( : ¡ G e V ) £ [ ph = c m = s ] Flux for TS > : = binUpper Limits -9 -10 -11 -12 -13 -14 -15 F l u x ( : ¡ G e V ) £ [ ph = c m = s ] Flux for TS > : = binUpper Limits2009 2010 2011 2012 2013 2014 2015 2016Year051015 F l u x ( : ¡ G e V ) £ [ ph = c m = s ] Flux for TS > : = binUpper Limits F l u x ( : ¡ G e V ) £ [ ph = c m = s ] Flux for TS > : = binUpper Limits F l u x ( : ¡ G e V ) £ [ ph = c m = s ] Flux for TS > : = binUpper Limits F l u x ( : ¡ G e V ) £ [ ph = c m = s ] Flux for TS > : = binUpper Limits Fig. 1. 5BZQJ 0359+5057 . Top left: TS map consideringonly E > γ -raysignature (used for all further TS maps), considering thebrightest period from 2014 to 2015; 5BZQJ 0359+5057 ismarked with a circle centered at “+”. Top right: The SEDwith a polynomial fit to the mean Syn and IC components;the blue-bump feature between 10 and 10 Hz (VandenBerk et al. 2001) for a source at z=1.512. In the γ -ray band,red points represent the SED before flaring (before 2013),while blue is for the flaring period (after 2013). Bottom:0.3-500 GeV light curve along 7.5 yrs of observations withbins of 30 days; red points represent upper limit flux.the 7.5 yrs covered by Fermi-LAT. Light curves are com-puted with likelihood analysis, therefore the background isextracted and flux points are calculated only for time binsthat have TS >
4. For all cases we compute upper limitsand errors bars using the integral method assuming a 95%confidence level, as provided by the Fermi Science Tools.When relevant for the discussion, we also show the signif-icance of the γ -ray signatures based on the 3FGL setup(i.e., likelihood analysis with 4.0 yrs of Pass 7 data) to testif those sources could have been identified previously. We present a detailed description for the new γ -ray detec-tion, which showed TS >
20 when integrating over 7.5years of observations. Those cases are relatively isolated,meaning there are no other close by γ -ray counterpartsthat could contaminate the observed signatures. For thediscussion we include TS maps, light curves, and spectralpoints. Together, those elements build an entire picture,not only validating and describing their γ -ray properties,but also explaining (whenever necessary) why those sourceswere not previously detected, showing examples on how thedata treatment is refined with multifrequency information,leading to more efficient use of public databases. Computing a TS map in the 0.5-12 GeV energy range (Fig. 1), we identified a bright point-like source clearly emerging from a low and flat TS back-ground. The light curve plotted in Fig. 1, shows that the F l u x ( : ¡ G e V ) £ [ ph = c m = s ] Flux for TS > : = binUpper Limits -9 -10 -11 -12 -13 -14 -15 -16 F l u x ( : ¡ G e V ) £ [ ph = c m = s ] Flux for TS > : = binUpper Limits2009 2010 2011 2012 2013 2014 2015 2016Year051015 F l u x ( : ¡ G e V ) £ [ ph = c m = s ] Flux for TS > : = binUpper Limits F l u x ( : ¡ G e V ) £ [ ph = c m = s ] Flux for TS > : = binUpper Limits F l u x ( : ¡ G e V ) £ [ ph = c m = s ] Flux for TS > : = binUpper Limits F l u x ( : ¡ G e V ) £ [ ph = c m = s ] Flux for TS > : = binUpper Limits
064 6.4 13 20 26 33 39 46 52 59 6 . . . . Right ascension D e c li na t i on +
064 6.4 13 20 26 33 39 46 52 59 6 . . . . Right ascension D e c li na t i on Fig. 2. BZQJ 1955+5131 . Top left: TS map consideringonly E > γ -ray spectrum in the 0.3-500 GeVband. As example, in the 10 -10 Hz band (0.1-500 GeV)we show the Fermi-LAT sensitivity limit when integratingover four year of observations. Bottom: The γ -ray lightcurve for 5BZQJ 1955+5131 along 7.5 yrs of observationsand with time bin of 30 days.source 5BZQJ 0359+5057 (4C +50.11) was undetectable byFermi-LAT, most of the time with the exception of the pe-riod between June 2013 and April 2016 when it underwenta phase of strong γ -ray activity. This is consistent with itsnondetection up to the 3FGL catalog (since 3FGL only in-tegrate observations from Aug. 2008 up to Aug. 2012).To reproduce the 3FGL setup, we performed alikelihood analysis with Pass 7 data, integrating onlyduring the first four years of observations. As re-sult we found TS P ass . yrs =15.2, with model parametersN = 8.1 ± × − ph/cm /s/MeV and Γ=2.55 ± =1000 MeV as pivot energy. Hence, this source wasout of the 3FGL catalog simply because it did not meet theTS > γ -ray signatures at 10 to 25 TS level contain rich informa-tion, as discussed in Arsioli & Chang (2017).As an exercise, we used a setup similar to 3FGL (inte-grating over 2008-2012) but now with PASS 8 data. Thisresults in a firm detection with TS >
25 and the γ -ray SEDshown as red points in Fig. 1 (bottom). This give us a solididea on how the Pass 8 data release brings way relevantenhancements for the description of the γ -ray sky. This blazar shows a strong γ -raysignal when integrating over 7.5 yrs with Pass 8 data, butfrom the light curve (Fig. 2) we measured significant flaringactivity only after 2014. A likelihood analysis integratingPass 7 data from Aug. 2008 up to Aug. 2012 (hereafter werefer to this as the 3FGL setup) have shown TS P ass . yrs ≈ ◦ ), where the low-energy dif-
5. Arsioli and G. Polenta: A complete sample of LSP blazars fully described in γ -rays .9 3.3 4.2 4.8 5.3 5.6 5.9 6.1 6.4 6.6 6. . . . . Right ascension D e c li na t i on .9 3.3 4.2 4.8 5.3 5.6 5.9 6.1 6.4 6.6 6. . . . . Right ascension D e c li na t i on .9 3.3 4.2 4.8 5.3 5.6 5.9 6.1 6.4 6.6 6. . . . . Right ascension D e c li na t i on + -9 -10 -11 -12 -13 -14 -15 -16 F l u x ( : ¡ G e V ) £ [ ph = c m = s ] Flux for TS > : = binUpper Limits2009 2010 2011 2012 2013 2014 2015 2016Year051015 F l u x ( : ¡ G e V ) £ [ ph = c m = s ] Flux for TS > : = binUpper Limits F l u x ( : ¡ G e V ) £ [ ph = c m = s ] Flux for TS > : = binUpper Limits F l u x ( : ¡ G e V ) £ [ ph = c m = s ] Flux for TS > : = binUpper Limits F l u x ( : ¡ G e V ) £ [ ph = c m = s ] Flux for TS > : = binUpper Limits F l u x ( : ¡ G e V ) £ [ ph = c m = s ] Flux for TS > : = binUpper Limits Fig. 3. 5BZUJ 0909+4253 . Top left: TS map consider-ing only E > γ -ray light curvealong 7.5 yrs of observations.fuse background is more intense thus hindering detectionsof faint γ -ray sources. In fact, owing to larger backgroundlevels, the Fermi-LAT sensitivity in the Galactic plane re-gion is lower than at high Galactic latitude . This blazar is strongly de-tected when integrating over 7.5 yrs with Pass 8 data(TS
P ass . yrs = 74.4), but shows no strong flaring activity ascan be seen in Fig. 3. A likelihood analysis with the3FGL setup gives TS P ass . yrs ≈ = 2.4 ± × − ph/cm /s/MeV, Γ = 3.04 ± = 1000 MeV as pivot energy, which agrees with val-ues reported in Table 1. This is another example for which itwould be beneficial to have preliminary information aboutfaint signatures with TS in between 10 to 25, which were al-ready available with 4 yr Pass 7 data. The TS map (Fig. 3)confirms that the observed γ -ray signature emerges as apoint-like source from a low-TS background. From the light curve (bottompanel of Fig. 4) this source shows γ -ray activity only in2014, which is consistent with its noninclusion in the 3FGLcatalog (3FGL covers the period of 08/2008-08/2012). Alikelihood analysis with 3FGL setup results in null detec-tion. For this case, we are likely probing the peak of atransient γ -ray activity during the year 2014, and smooth-ing out the signal along 7.5 years of binned analysis. Thedetection of a steady high-energy component (non-flaringstate) is currently limited by the Fermi-LAT sensitivity.Therefore, the parameters reported on Table 1 embody themean γ -ray spectrum behavior and represent a good exam-ple of how γ -ray activity is washed out for the building ofcurrent high-energy catalogs that integrate Fermi-LAT ob-servations over 4.0 year (in case of the 3FGL) to 7.5 years.We built a TS map (top left in Fig. 4) integrating along For a description on the Fermi-LAT performance check . . . Right ascension D e c li na t i on . . . Right ascension D e c li na t i on + -9 -10 -11 -12 -13 -14 -15 -16 . . . Right ascension D e c li na t i on F l u x ( : ¡ G e V ) £ [ ph = c m = s ] Flux for TS > : = binUpper Limits F l u x ( : ¡ G e V ) £ [ ph = c m = s ] Flux for TS > : = binUpper Limits2009 2010 2011 2012 2013 2014 2015 2016Year051015 F l u x ( : ¡ G e V ) £ [ ph = c m = s ] Flux for TS > : = binUpper Limits F l u x ( : ¡ G e V ) £ [ ph = c m = s ] Flux for TS > : = binUpper Limits F l u x ( : ¡ G e V ) £ [ ph = c m = s ] Flux for TS > : = binUpper Limits F l u x ( : ¡ G e V ) £ [ ph = c m = s ] Flux for TS > : = binUpper Limits Fig. 4. 5BZQJ 1153+8058 . Top left: TS map usingE > - 10 Hz assuming z=1.250. Bottom: The γ -raylight curve for 5BZQJ 1153+8058 along 7.5 yrs.the whole year of 2014, showing that the blazar position iscompatible with the γ -ray signature within the 68% con-tainment region. This case in particular allowed us to useE > >
500 MeV, providing im-proved localization.
Although this source is locatedin a relatively crowded region close to the Galactic plane( | b | =17.9 ◦ ), the low-energy detection is very significantwith TS P ass . yrs ≈
80 in the 300-550 MeV energy bin alone.We found no significant flaring activity during the 7.5 yrsof observations with Fermi-LAT, and the period with themost significant γ -ray signature extends from October 25 toNovember 30, 2010. The highest energy photons detectedfrom this region are ∼
10 GeV, and therefore we built a TSmap in the 500 MeV - 12 GeV energy range (Fig. 5, leftpanel). The γ -ray signature emerges as a point source fullycompatible with the blazar position. In particular, the γ -raycounterpart has high redshift of z = 3.396, and absorptiondue to EBL might hinder the detection of VHE photons.A likelihood analysis using the 3FGL setup hasshown TS P ass yrs = 20.3, with pivot energy E = 1000 MeV,pre-factor N = 2.6 × − ph/cm /s/MeV, and Γ = 3.61,which is in good agreement with the parameters presentedin Table 1. This is another example of low-significance γ -ray signature with TS between 10 and 25, which would havebeen beneficial to report on without compromising the spec-tral description (that can be refined with longer integrationtime as shown). This source has been detectedwith TS ≈
20 when integrating over 7.5 yrs with Pass 8 data.Its light curve does not present strong flaring episodes, butwe identify the most relevant bin as July 31 to August 31,2013, as used to build the TS map. Although the TS value isbelow 25, we consider it a firm detection since the TS map(6) clearly shows the γ -ray point-like signature emerge froma low-TS background, and the 68% confinement region is
6. Arsioli and G. Polenta: A complete sample of LSP blazars fully described in γ -rays
032 0.85 1.7 2.6 3.4 4.3 5.1 6 6.8 7.7 8. . . . . Right ascension D e c li na t i on +
032 0.85 1.7 2.6 3.4 4.3 5.1 6 6.8 7.7 8. . . . . Right ascension D e c li na t i on -12 -13 -14 -15 -16 Fig. 5. 5BZQJ 0646+4451 . Left: TS map considering500 MeV to 12 GeV photons and integrating over the flar-ing period from October 25 to November 30, 2010. Blackdashed lines represent 68% and 95% confinement regionsfor the γ -ray signature. Right: SED for 5BZQJ 0646+4451,also showing the blue bump template assuming z=3.396. . . - . Right ascension D e c li na t i on + . . - . Right ascension D e c li na t i on -9 -10 -11 -12 -13 -14 -15 -16 . . - . Right ascension D e c li na t i on Fig. 6. 5BZQJ 0125-0005 . Left: TS map considering onlyE >
500 MeV photons and integrating data along thebrightest month from July 31 to August 31, 2013. Blackdashed lines show the 68% and 95% confinement regionfor the γ -ray signatures. Right: Multifrequency SED for5BZQJ 1153+8058 with the blue bump template corre-sponding to z=1.077compatible with the blazar position. A likelihood analysiswith the 3FGL setup gives TS P ass yrs = 12.9 with parame-ters E = 1000 MeV, N = 2.8 × − ph/cm /s/MeV, andΓ = 2.68, again in agreement with parameters from Table 1. We discuss a case in which source confusion involving asteep and a hard spectrum γ -ray source is likely present.Even though no γ -ray source is reported within 1 ◦ of5BZQJ 1642+6856 in any of the FGL catalogs, our like-lihood analysis finds a γ -ray signature matching the po-sition of this source, which is located only ∼
10 arcminfrom the HSP blazar 2WHSPJ 164014.8+685233. Indeed,the high-energy E > γ -ray signature coincident withthe 2WHSP source, however extending toward the posi-tion of 5BZQJ 1642+6856. We then consider photons in the0.1-500 GeV band and a likelihood function including twonearby point-like emitters, which gives us an estimate ofthe γ -ray spectral properties for each source as shown inTable 2. Despite the low statistical significance associatedwith BZQJ 1642+6856 γ -ray signature (TS < > γ -raysignature is now part of the background model. As a re-sult (Fig. 7, top right panel) we reveal a residual point-like signature consistent with 5BZQJ 1642+6856 within the68% confinement radius. All together, it suggests this mightbe a case of source confusion, which is hard to resolvewith currently available data. Since those γ -ray signatureshave not been reported in previous high-energy catalogs, wepresent 2WHSPJ 164014.8+685233 as a new detection and5BZQJ 1642+6856 as a relevant signature, whose resolvedSEDs are shown in Fig. 7 (middle and bottom panels). Table 2.
Source model parameters derived from the FermiScience Tools assuming a power-law to describe the γ -rayspectrum within the 0.1-500 GeV energy band, with N given in ph/cm /s/MeV, and assuming E = 1000 MeV. Source N (10 − ) Γ TS2WHSPJ 164014.8+685233 7.4 ± ± ± ± Arsioli & Chang (2017) showed that high-energy TS mapscan be used to improve the localization for many γ -ray sig-natures currently listed in FGL catalogs. It is well knownthat the Fermi-LAT detector (Atwood et al. 2009) is charac-terized by a highly energy-dependent point spread function(PSF), which contains 68% of the 1 GeV events within 0.8 ◦ ,decreasing afterward with a trend ∝ E − . up 10’s GeV,and finally roughly constant at 0.1 ◦ up to the highest en-ergies considered in this paper. Therefore, working withE > γ -ray signature, which helps solve cases ofsource confusion. This is particularly important for unas-sociated 3FGL (Acero et al. 2013) that are actually coun-terparts of close by blazars. We present two such cases,noting that an approach based on the prior multifrequencyidentification of nearby blazars certainly improve the γ -rayassociations with potential counterparts. We present an improved posi-tion reconstruction for the source 3FGLJ 0432.5+0539, forwhich no association is reported on the 3FGL catalog. Webuild a TS map using only photons with E > γ -ray signature is fully consistent with the position ofBZUJ 0433+052, while the 3FGL position (magenta dashedcircle on the top left panel) is well outside of the 99% con-finement region for the high-energy TS peak. Although partof the improvement could be ascribed to the better instru-ment response function (IRF) and event selection of thePass 8 data release with respect to the Pass 7 as used for3FGL production, it should be noted that high-energy TSmaps has proven to provide a significant contribution insource positioning. This is another case of a 3FGLsource with no association that benefits from an improvedposition reconstruction. As shown in the top right panel of
7. Arsioli and G. Polenta: A complete sample of LSP blazars fully described in γ -rays -11 -12 -13 -14 -15 -16 -11 -12 -13 -14 -15 e-05 3.6 7.1 11 14 18 21 25 29 32 3 . . . Right ascension D e c li na t i on e-05 3.6 7.1 11 14 18 21 25 29 32 3 . . . Right ascension D e c li na t i on . . . Right ascension D e c li na t i on . . . Right ascension D e c li na t i on Pure TS map: E > 1 GeV Residual: E > 1 GeV 0 0.4 2.1 + + + +
Fig. 7. BZQJ 1642+6856 . Top left: TS map forE > > γ -ray background; in this case the blackdashed line corresponds to the 68% confinement radius forthe γ -ray source, while the white cross indicates the po-sition of BZQJ 1642+6856. Middle and bottom: SEDs for2WHSPJ 1640+6852 for which no optical identification isavailable yet, and BZQJ 1642+6856 with z = 0.751. The up-per limits on the γ -ray spectra have been calculated onlyfor energy bins with low significance.Fig. 9, there is no relevant radio or X-ray counterpart thatis compatible with the 95% positional error ellipse for the3FGL source (dot-dashed line). However, a powerful blazaris only ∼
18 arcmin away. Indeed, building a TS map usinghigh-energy photons (with E > γ -ray signatureis fully consistent with the position of 5BZQJ 0555+3948.In addition, the γ -ray spectrum we obtained is compatiblewith expectations for the end tail of the IC bump. Blazar 5BZQJ 0228+6721 had no γ -ray counterpart in pre-vious FGL catalogs, however a γ -ray source at its vicinity − −
10 0 10 20 − − arcmin a r c m i n BZUJ0433+0521
12 2.8 5.7 8.6 11 14 17 20 23 26 2 . . . Right ascension D e c li na t i on
12 2.8 5.7 8.6 11 14 17 20 23 26 2 . . . Right ascension D e c li na t i on + − −
10 0 10 20 − − arcmin a r c m i n BZUJ0433+0521
12 2.8 5.7 8.6 11 14 17 20 23 26 2 . . . Right ascension D e c li na t i on
12 2.8 5.7 8.6 11 14 17 20 23 26 2 . . . Right ascension D e c li na t i on + -9 -10 -11 -12 -13 -14 -15 Fig. 8. 5BZUJ 0433+0521 . Top left: TS map built usingonly photons with E > ∼ o ) and the intense low-energy diffuse γ -raybackground can hinder both the source localization and de-tection when integrating over the full energy bandwidth.Since the 3FGL position is only 5.94 arcmin from theblazar, we investigated whether we could improve the γ -ray localization of this source working with E >
500 MeVphotons, benefiting from lower background intensity, im-proved PSF, and longer integration time (from 4.0 to 7.5years). We calculated the E >
500 MeV TS map (Fig. 10,top left) showing that the 68% containment region for the γ -ray signature is compatible with 5BZQJ 0228+6721 (whileGB6J 0229+6706 is out of the 99% containment region) andtherefore the 3FGL association should be revised. We recal-culated the γ -ray SED (Fig. 10, bottom) assuming a single
8. Arsioli and G. Polenta: A complete sample of LSP blazars fully described in γ -rays − −
10 0 10 20 − − arcmin a r c m i n
099 13 21 26 30 33 36 38 40 41 4 . . . Right ascension D e c li na t i on
099 13 21 26 30 33 36 38 40 41 4 . . . Right ascension D e c li na t i on + -9 -10 -11 -12 -13 -14 -15 -16 Fig. 9. 5BZQJ 0555+3948 . Top right: Sky-Explorerview, showing 3FGLJ 0556.2+3933 indicated as a red cross,while 5BZQJ 0555+3948 is outside the error circle associ-ated with the γ -ray signature from 3FGL database. Blueand red circles represent X-ray and radio detections inthis region. Top left: TS map, E > γ -ray signature; Black dashed lines showthe 68%, 95%, and 99% confinement region for the γ -raysignatures. The dashed magenta circle corresponds to the3FGLJ 0556.2+3933 position, while the green circle is cen-tered on the position of 5BZQJ 0555+3948 marked as blackcross. Bottom: SED for 5BZQJ 0555+3948, where the bluebump template assumes z=2.365.source with position corresponding to 5BZQJ 0228+6721,and found it is consistent with the end-tail IC bump. γ -ray excesses We present sources showing faint γ -ray signature10 < TS <
20 when integrating over 7.5 years of Fermi-LATobservations. It is important to report on faint detections,especially to clarify if these sources are actually γ -ray ac-tive, but under the TS limit currently used by the Fermiteam, or if they are really quiet in the γ -ray band. Also, low-significance γ -ray detections help to complement the sourcenumber count in the faint end of the logN-logS γ , and there-fore can impact estimates of the contribution of blazars tothe E < γ -ray background. We foundthe light curves for 5BZQJ 2218 − − − −
10 0 10 20 − − arcmin a r c m i n GB6 J0229+6706
45 2.2 4.8 7.5 10 13 15 18 21 23 2 . . . . Right ascension D e c li na t i on
45 2.2 4.8 7.5 10 13 15 18 21 23 2 . . . . Right ascension D e c li na t i on GB6 J0229+6706 + X X -9 -10 -11 -12 -13 -14 -15 -16 Fig. 10. 5BZQJ 0228+6721 . Top left: Sky Explorer view,showing 3FGLJ 0228.5+6703 marked with “x” symbol, anddashed line for the error circle associated with the 3FGL γ -ray signature. We also highlight the positions of GB6J0229+6706 and 5BZQJ 0228+6721. The red and blue cir-cles represent radio and X-ray detections in the field. Topright: TS map considering E >
500 MeV photons, withdashed lines representing 68%, 95%, and 99% containmentregion for the γ -ray signature; 5BZQJ 0228+6721 is centerat “+” matching the TS peak position. Bottom: SED for5BZQJ 0228+6721, with blue bump assuming z=0.523. Table 3.
Source model parameters from Fermi ScienceTools, assuming a power-law to describe the γ -ray spectrumwithin 0.1-500 GeV, with N given in [ph/cm /s/MeV], andusing E = 1000 MeV as pivot energy. Case marked with *is probably an spurious detection. Source z N (10 − ) Γ TS5BZQJ 2218 − ± ± − ± ± ± ± ± ± As seen from the TS maps, only 5BZQJ 1927+7358 isout of the 68% containment for the γ -ray signature. In thiscase, the observed TS value (table 3) when integrating over7.5 yrs of observations can be attributed to residual signalfrom an unidentified close-by γ -ray source. For now, weconsider it as a likely spurious signal.
9. Arsioli and G. Polenta: A complete sample of LSP blazars fully described in γ -rays
56 1.6 2.7 3.8 4.9 6 7 8.1 9.2 10 1 . . Right ascension D e c li na t i on
046 0.66 1.3 1.9 2.5 3.1 3.7 4.3 4.9 5.6 6. . . . . Right ascension D e c li na t i on
002 2 3.9 5.9 7.8 9.8 12 14 16 18 2 - . - . - . Right ascension D e c li na t i on - . - . - . Right ascension D e c li na t i on .6 1.5 2.3 3.2 4.1 4.9 5.8 6.6 7.5 8.4 9. . . Right ascension D e c li na t i on
002 2 3.9 5.9 7.8 9.8 12 14 16 18 2 - . - . - . Right ascension D e c li na t i on - . - . - . Right ascension D e c li na t i on + + + +
12 0.6 1.1 1.6 2 2.5 3 3.5 4 4.5 4. . . . . Right ascension D e c li na t i on Fig. 11.
TS maps for the low-significance γ -ray detectionsusing only E >
500 MeV photons. For each case the blazarposition is highlighted with thick green circle center at “+”.Black dashed lines show the 50%, 68%, and 95% contain-ment radius for the γ -ray signature. We present the γ -ray analysis for three radio-loud blazars that have relatively bright ν f ν Synpeak, 5BZQJ 0010+1058, 5BZUJ 0241-0815, and5BZQJ 2136+0041, for which the detection with Fermi-LAT was expected but not yet reported. This brings us toan important consideration about the role of short-livedflares at monthly timescales since a considerable portionof γ -ray active sources could be still undetected simplybecause their short-lived signatures are diluted belowthe sensitivity threshold when integrating Fermi-LATobservations over long exposure time. This object is a radio-loud source(also known as MRK 1501) for which γ -ray detectionwas expected since it is a relatively bright and close-byblazar with ν f ν =10 − . erg/cm /s and z=0.089. When in-tegrating over 7.5 years, no γ -ray signature was evident(TS ≈ γ -ray spectrum, reaching TS ≈
26 forthe single month bin of June 2010. In this case (Fig. 12,top right), the upper limits calculated for the γ -ray SED areless restrictive, since we are integrating over a single month.We checked for coincident X-ray measurements from Swift (along the γ -ray flaring activity from June 2009 to May2010), but for the observations made in February 2010 wecould find no sign of strong X-ray flaring. This is a goodexample of transient γ -ray source, which we could only de-
012 1.5 2.9 4.4 5.8 7.3 8.7 10 12 13 1 . . . Right ascension D e c li na t i on
012 1.5 2.9 4.4 5.8 7.3 8.7 10 12 13 1 . . . Right ascension D e c li na t i on + F l u x £ [ ph = c m = s ] Flux for TS > : = binUpper Limits -9 -10 -11 -12 -13 -14 -15 -16 F l u x £ [ ph = c m = s ] Flux for TS > : = binUpper Limits Fig. 12. 5BZQJ0010+1058 . Top left: TS map consider-ing E >
500 MeV photons during a short-flare episode atMET: 296164808-298860398. Black dashed lines represent-ing 68% and 95% containment region for the γ -ray sig-nature. Top right: The corresponding SED of this object.Bottom: Light curve for 5BZQJ 0010+1058 background ex-tracted with likelihood analysis; flux points are calculatedonly for bins having TS > γ -ray candidates. NGC1052; 5BZUJ 0241-0815.
In this case, the lightcurve shows a flaring state during 2013 to first quarterof 2014. From March 7 to April 7, 2014 the γ -ray signa-ture reaches its highest state, at which TS ≈
12. A TS mapat E >
500 MeV (fig. 13) shows that 5BZUJ 0241-0815 iswithin the 68% confinement region for the γ -ray signature.This blazar of unknown type is another example of a tran-sient γ -ray emitter, which may contribute to building thecurrently unresolved γ -ray background. It is still not clearhow to evaluate the impact of such sources for the observedextragalactic diffuse γ -ray component given that transientpopulations are not yet fully characterized. This source is classified as aFSRQ with z=1.941, and for this case we detected a short-lived γ -ray flare from January 2 to February 2, 2014, whereTS ≈ >
300 MeV photons(Fig. 14, top left) shows that the signature emerges as apoint-like source and the 68% containment region is com-patible with 5BZQ J2136+0041 position. Although the γ -ray signature has low significance, it is important to reportit as a potential faint γ -ray transient. Since we integratedalong short time period, the lower limits computed in the γ -ray SED (Fig. 14, top right) are not so restrictive.From the three cases presented in this section, two ques-tions clearly arise: How many AGNs could have short-livedflares, and how important is their integrated contributionto the γ -ray background probed by Fermi-LAT (Ackermannet al. 2015a) in the 100 MeV to 800 GeV energy range?
10. Arsioli and G. Polenta: A complete sample of LSP blazars fully described in γ -rays - . - . Right ascension D e c li na t i on - . - . Right ascension D e c li na t i on + : : : : : F l u x £ [ ph = c m = s ] Flux for TS > : = binUpper Limits : : : : : F l u x £ [ ph = c m = s ] Flux for TS > : = binUpper Limits : : : : : F l u x £ [ ph = c m = s ] Flux for TS > : = binUpper Limits -8 -9 -10 -11 -12 -13 -14 -15 -16 Fig. 13. 5BZUJ 0241-0815 . Top left: TS map forE >
500 MeV during a short-flare episode at MET:415846402-418598890. The blazar position is highlighted bythe green circle center at “+”; the black dashed-lines are TSsurfaces corresponding 68%, 95%, and 99% confinement re-gion for the γ -ray signature. Top right: The correspondingSED of this object; the green template corresponds to theelliptical galaxy emission at z=0.005. Bottom: Light curvefor 5BZU J0241-0815. Flux points are only calculated forbins with TS > We report on four Radio-Planck sources (5BZQJ0927+3902, 5BZQJ 2139+1423, 5BZQJ 2022+6136 and5BZQJ 2007+4029) for which we could not find evidenceof γ -ray signature during the 7.5 years of observations. This is a bright radio blazar(z = 0.695) about 46.1 arcminutes away from 3FGLJ0923.1+3853 (which is associated to B2 0920+39). Weinvestigate this region with a TS map, looking for signsof source confusion, but we could find none. The 3FGL γ -ray signature dominates the emission in this region asseen by the low-energy TS map (Fig. 15, left); the SEDfor 5BZQJ 0927+3902 (Fig. 15, right) has no of γ -ray in-formation. This is likely a good proxy for blazars withan IC component that is MeV peaked just as the fol-lowing cases: 5BZQJ 2139+1423, 5BZQJ 2022+6136, and5BZQJ 2007+4029. This blazar is classified as FSRQ,with relatively high redshift, z = 2.427. A light curve withone month bins along 7.5 year showed no bins withTS > γ -ray activity forthis source. The IC component probably peaks at a fre-quency that is much lower than the bandwidth probed byFermi-LAT, owing to its LSP frequency log( ν peak ) ≈ This source is also classified asFSRQ, although there is no good fitting between the optical
019 0.59 1.2 1.8 2.4 3 3.5 4.1 4.7 5.3 5. . . Right ascension D e c li na t i on
019 0.59 1.2 1.8 2.4 3 3.5 4.1 4.7 5.3 5. . . Right ascension D e c li na t i on + F l u x £ [ ph = c m = s ] Flux for TS > : = binUpper Limits F l u x £ [ ph = c m = s ] Flux for TS > : = binUpper Limits F l u x £ [ ph = c m = s ] Flux for TS > : = binUpper Limits -9 -10 -11 -12 -13 -14 -15 -16 Fig. 14. 5BZQJ 2136+0041 . Top left: TS map forE > γ -ray signature. Top right: The corresponding SED of thisobject; the blue bump template assumes z=1.941. Bottom:Light curve for 5BZQJ2136+0041, flux is calculated onlyfor TS >
045 3.9 7.8 12 16 20 24 28 31 35 3 . . . Right ascension D e c li na t i on . . .
045 3.9 7.8 12 16 20 24 28 31 35 3 . . . Right ascension D e c li na t i on
045 3.9 7.8 12 16 20 24 28 31 35 3 . . . Right ascension D e c li na t i on
045 3.9 7.8 12 16 20 24 28 31 35 3 . . . Right ascension D e c li na t i on -9 -10 -11 -12 -13 -14 -15 -16 Fig. 15. 5BZQJ 0927+3902 . Left: TS map for 750-950MeV photons. The blazar 5BZQJ 0927+3902 is highlightedas a green circle and 3FGL is shown in magenta. Right:The SED with a polynomial fit to the mean Syn and theblue-bump template are represented for z=0.695. In the γ -ray band, the blue curve in the 10 -10 Hz band (0.1-500 GeV) represents the Fermi-LAT four year sensitivitythreshold, therefore an upper limit for the γ -ray emission.to X-ray data and a blue-bump template (with z = 0.228).There is a single episode from 27 September to 28 October2011 for which a low significance γ -ray signal (TS = 4.1) ispresent. This object is close to theGalactic plane, at latitude b = 4.30 ◦ , with relatively highredshift, i.e., z = 1.736. A likelihood analysis integrat-ing over 7.5 years, considering the full energy band 0.1-500 GeV shows a very low-significance γ -ray signature,where TS ≈ ± = 2.4 ± × ph/cm /s/MeV. However, from the TS map (300 MeV-10 GeV) there is no clear evidence for a point source, there-fore we do not consider this as detection.
11. Arsioli and G. Polenta: A complete sample of LSP blazars fully described in γ -rays : : : : : : : : º Syn : peak )0510152025 N u m b e r o f s o u r ce s Histogram of Log( º Syn : peak )Median value : 12 : : : : : : : : : º Syn : peak )0510152025 N u m b e r o f s o u r ce s Histogram of Log( º Syn : peak )Median value : 12 : : : : : : : : : º Syn : peak )0510152025 N u m b e r o f s o u r ce s Histogram of Log( º Syn : peak )Median value : 12 : Mean Value: 12.94
Fig. 16.
Distribution of log( ν Synpeak ) [Hz] for the Radio-Planck sample considering all 104 objects, bin size of 0.3;the mean value of 10 . Hz is shown as a dashed line.
5. Radio-Planck sample properties
In previous sections, we showed that 99 of the 104 objects inthe Radio-Planck sample have evidence for γ -ray emissionat relevant level when integrating over 7.5 years of obser-vation or during flaring episodes. By fitting the Syn and ICcomponents with a third order polynomial (Giommi et al.2012b), we estimate their SED peak parameters (Table 5)using all available nonsimultaneous data. In particular, the γ -ray points are from the 3FGL catalog in the case of pre-viously detected sources, and from our own data reductionin the case of newly detected sources, considering 7.5 yearsof Fermi-LAT observations; we also study statistical prop-erties related to the nonthermal peak parameters. In Fig. 16 we plot the distribution of log( ν Synpeak ), showingthat the Radio-Planck sample is dominated by LSP blazars,with mean value (cid:104) log( ν Synpeak ) (cid:105) = 12.94 ± [Hz]. This isbecause our sample is flux limited in the microwave bandwhere blazars of the LSP type are by far the most abundantobjects.In Fig. 17 we plot a histogram showing the distribu-tion of log( ν f ν ) peak values for both Syn and IC bumps.A parametric Kolmogorov Smirnov (KS) test comparingboth histograms gives a p value = 0.86, implying that the lu-minosity distributions of both components are similar. Infact, the mean values of the peak fluxes are relatively close,i.e., (cid:104) log( ν f Synν ) (cid:105) = -11.11 ± /s]; (cid:104) log( ν f ICν ) (cid:105) = -10.94 ± /s]. This similarity in the peak powerdistribution of the two components suggests that on aver-age the ratio of ν f Synν to ν f ICν values might be close to onefor the population of LSP blazars.
The ν Synpeak and ν ICpeak parameters, which are determinedin the log( ν f ν ) versus log( ν ) plane, are very representa-tive of blazar SED properties because they give the peakenergy where most of the Syn and IC power are emit-ted. To study their statistical properties we define the From here on, the errors to the mean values are calculatedas σ/ √ n , where σ is the standard deviation and n is the totalnumber of objects in each sample (or subsample). ¡ : ¡ : ¡ : ¡ : ¡ : ¡ : ¡ : ¡ : º f º ) [erg = cm = s]051015202530 N u m b e r o f S o u r ce s Synchrotron peakIC peak ¡ : ¡ : ¡ : ¡ : ¡ : ¡ : ¡ : ¡ : º f º ) [erg = cm = s]051015202530 N u m b e r o f S o u r ce s Synchrotron peakIC peak ¡ : ¡ : ¡ : ¡ : ¡ : ¡ : ¡ : ¡ : º f º ) [erg = cm = s]051015202530 N u m b e r o f S o u r ce s Synchrotron peakIC peak : : : : : : : : º Synpeak )01020 23 : £ exp ¡ ¡ (x ¡ : )0 : ¢ ¡ : ¡ : ¡ : ¡ : ¡ : ¡ : ¡ : ¡ : º f º ) [erg = cm = s]051015202530 N u m b e r o f S o u r ce s Synchrotron peakIC peak
Fig. 17.
Histogram of log( ν f ν ) at the Syn peak (in red)and the IC peak (in blue), considering all 99 sources in theRadio-Planck sample for which we could estimate the Synand IC peak parameters.peak ratio (PR) parameter as the logarithm of ν peak ra-tios: PR = log( ν ICpeak /ν Synpeak ) and plot its distribution (Fig.18). We determine the characteristic mean value for PRwhen considering all LSP blazars: (cid:104) PR all (cid:105) =8.60 ± (cid:104) PR BL Lacs (cid:105) =8.42 ± (cid:104) PR F SRQ (cid:105) =8.75 ± ν ICpeak ) distribution for all 98LSPs with available IC data (in blue), which has mean valueof (cid:104) log( ν IC ) (cid:105) =21.53 ± (cid:104) PR all (cid:105) value to estimate ν ICpeak based on the ν Synpeak according tolog( ν ICpeak ) = log( ν Synpeak )+ (cid:104) PR all (cid:105) . The distribution of ν ICpeak calculated via (cid:104) PR all (cid:105) parameter is shown in pink, which isindeed well described by this simple relation. Most likely,there is a dominant process connecting Syn and IC bumps,otherwise such correlations would not show up. If multi-ple emission scenarios were at work, we would expect largespreading in the parameter space, not tight Gaussian distri-butions. As known, the most established picture to describethe SED shape assumes dominant SSC leptonic scenario,but there is extensive discussion in the literature consider-ing the role of the EC for different blazars, even reportingon observable evidence (Meyer et al. 2012).A scatter plot with log( ν IC ) values versuslog( ν Syn )+ (cid:104) PR all (cid:105) shows that there is only marginalevidence for the correlation between those parameter(Fig. 19, bottom) given that the Pearson’s correlationcoefficient r for a linear fit is ∼ ν IC )as a population, tells us that the log( ν IC ) measured by ourfitting might have large uncertainties; given the absenceof a significant correlation, when comparing case by casewith the scatter plot. Indeed there is a large uncertaintyfor that parameter mainly due to the huge data gap in theenergy window between tens of KeV up to hundreds ofMeV, which is smoothed out when considering the wholepopulation of bright LSP sources.
12. Arsioli and G. Polenta: A complete sample of LSP blazars fully described in γ -rays º IC º Syn )]00 : : : : N o r m a li ze dnu m b e r All Radio ¡ Planck5BZBz (BL Lacs)5BZQs (FSRQ) º IC º Syn )]00 : : : : N o r m a li ze dnu m b e r All Radio ¡ Planck5BZBz (BL Lacs)5BZQs (FSRQ) º IC º Syn )]00 : : : : N o r m a li ze dnu m b e r All Radio ¡ Planck5BZBz (BL Lacs)5BZQs (FSRQ)
5 6
7 8
9 10 Fig. 18.
Peak ratio log( ν ICpeak /ν Synpeak ) distribution. Full in-digo bars represent the whole Radio-Planck sample, reddashed bars indicated the subsample of BL Lacs, and greendot-dashed bars show the subsample of FSRQ.
18 19 20 21 22 23 24 25Log( º IC ) [Hz]05101520 N u m b e r o f s o u r ce s Log( º IC )Log( º Syn ) + PR
18 19 20 21 22 23 24 25Log( º IC ) [Hz]2021222324 L og ( º S y n ) + h P R a ll i [ H z ] Linear ¯t
18 19 20 21 22 23 24 25Log( º IC ) [Hz]2021222324 L og ( º S y n ) + h P R a ll i [ H z ] Linear ¯t
18 19 20 21 22 23 24 25Log( º IC ) [Hz]2021222324 L og ( º S y n ) + h P R a ll i [ H z ] Linear ¯t
Fig. 19.
Top: Distribution of log( ν ICpeak ) in blue, usingmeasured values from fitting the IC component, and inpink using log( ν ICpeak ) = log( ν Synpeak )+ (cid:104) PR all (cid:105) to estimate ν IC based on ν Syn measurements. Bottom: A scatter plot withlog( ν IC ) values vs. log( ν Syn )+ (cid:104) PR all (cid:105) together with a lin-ear fit showing week correlation. The ratio of IC to Syn peak power is known as the Comptondominance (CD). This is an important parameter for de-scribing blazar SEDs (Finke 2013; Potter & Cotter 2013;Nalewajko & Gupta 2017), since it measures the dominantpower output component for each source, i.e.,CD = L IC L Syn ≡ ν f IC ν ν f Syn ν ; at the SED peaks . (2)Luminosity is written as L = 4 π d L ν f ν /(1+z) − α , whered L represents the luminosity distance and the (1+z) − α factor is the k-correction assuming a power-law spectrumwith energy index α . Since we calculate the CD using the ¡ : ¡ : : : : N u m b e r o f s o u r ce s Radio Planck Sample5BZB (BL Lacs)5BZQ (FSRQ)5BZU (Uncertain) ¡ : ¡ : : : : N u m b e r o f s o u r ce s Radio Planck Sample5BZB (BL Lacs)5BZQ (FSRQ)5BZU (Uncertain) ¡ : ¡ : : : : N u m b e r o f s o u r ce s Radio Planck Sample5BZB (BL Lacs)5BZQ (FSRQ)5BZU (Uncertain) ¡ : ¡ : : : : N u m b e r o f s o u r ce s Radio Planck Sample5BZB (BL Lacs)5BZQ (FSRQ)5BZU (Uncertain) ¡ : ¡ : : : : N u m b e r o f s o u r ce s Radio Planck Sample5BZB (BL Lacs)5BZQ (FSRQ)5BZU (Uncertain) ¡ : ¡ : : : : N u m b e r o f s o u r ce s Gaussian ¯t : 24 £ exp( ¡ (x ¡ : : )Median at Log(CD) = 0 : Fig. 20.
Log(CD) distribution for the Radio-Planck sam-ple. Top: The green bars represent all 99 cases that haveSyn + IC data available for calculating the CD param-eter. The red dashed line represents a Gaussian functionwith σ =0.22 around the median value of log(CD) = 0.1.Bottom: The log(CD) distribution for BL Lacs (red dashedlines), FSRQ (blue dot-dashed lines), and blazars classifiedas uncertain are shown in green dashed lines.luminosity ratio at the peak power, α =1 for both IC andSyn peaks, and the luminosity ratio is simply the flux ratio.In Fig. 20 (top) we plot in green the distribution oflog(CD) parameter for the Radio-Planck sample, which hasa median value of 0.1. The median is only slightly largerthan 0, implying that on average, the peak-power output forthe sync and IC components are similar. In Fig. 20 (top),we add a tentative Gaussian fit to the log(CD) distribu-tions, showing that a single Gaussian function (red dashedline) hardly describes the overall shape, particularly thehighest CD values. The tail toward log(CD) > γ -rays compared to radiobands, pushing log(CD) to high values. The mean value (cid:104) log(CD) (cid:105) = 0.17 ± γ -ray activity in LSP blazars.The histograms at bottom panel of Fig. 20 representthe CD for different subsamples defined according to the5BZcat classification, that is, BL Lacs, FSRQ, and unclassi-fied sources. Nalewajko & Gupta (2017) have also estimatedthe CD for FSRQ and BL Lacs, but based on luminositiesat fixed energies: L (1 GeV) as measured with Fermi-LAT,and L (3 . µ m) as measured with the W1 channel from WISEsatellite, for instance, CD = L (1 GeV) / L (3 . µ m) . Our mea-surements instead are taken at the peak of Syn and IC com-ponents. In both cases, there is a trend for BL Lacs to pop-ulate the log(CD) range with the lowest values (with mean (cid:104) log(CD) (BL − Lac) (cid:105) = − . ± . (cid:104) log(CD) (FSRQ) (cid:105) = 0 . ± .
13. Arsioli and G. Polenta: A complete sample of LSP blazars fully described in γ -rays log(CD) distribution depends on radio flux density splittingthe sample in two subsets with f > ≤ IC , inMeV). When considering the whole sample, there is no clearcorrelation between those parameters (the Pearson’s corre-lation coefficient r ≈ ¡ ¡ IC ) [MeV] ¡ : ¡ : : : : : L og ( C o m p t o n D o m i n a n ce ) BZU ¡ UncertainBZQ ¡ FSRQBZB ¡ BL Lac ¡ ¡ IC ) [MeV] ¡ : ¡ : : : : : L og ( C o m p t o n D o m i n a n ce ) BZU ¡ UncertainBZQ ¡ FSRQBZB ¡ BL Lac ¡ ¡ IC ) [MeV] ¡ : ¡ : : : : : L og ( C o m p t o n D o m i n a n ce ) BZU ¡ UncertainBZQ ¡ FSRQBZB ¡ BL Lac
Fig. 21.
Compton dominance vs. energy associated withthe IC peak (E IC ) for the Radio-Planck sample. We plotFSRQs as blue dots, BL Lacs as red dots, and uncertainblazars as green crosses. The red dotted line sets a qualita-tive cut to highlight the region populated by BL Lacs. According to Acero et al. (2015) the variability index indi-cates if a γ -ray source is variable on a timescale of months,not addressing shorter or longer time variations. An in-dex > >
99% confidence probability thatthe source is variable. At least ≈ > γ -ray log(Var . index) takenfrom the 3FGL catalog versus our estimate of the param-eter log(CD). A linear fit of the scatter plot (Fig. 22) hasa Pearson correlation coefficient of 0.40, meaning the posi-tive correlation between log(Var . index) and log(CD) is rel-atively weak. In fact, this only tells us that the variabilityindex might not be the best parameter to rely on if we arewilling to investigate the influence of γ -ray variability overthe CD parameter.Looking at individual cases however provides a betterpicture of CD variations induced by fast variability. Weconsider first the three sources detected in γ -rays only dur-ing flaring episodes (Sec. 4.5). These objects move fromlog(CD) < − γ -ray quiet period) up to 0.79, 0.31,and 2.51, respectively, for BZQJ 0010+1058, BZUJ 0241- ¡ : : : : L og ( V a r : I nd e x )
5 4 3 ¡ : : : : L og ( V a r : I nd e x ) -0.5 ¡ : : : : L og ( V a r : I nd e x ) Fig. 22. γ -ray variability index vs. log(CD) for the Radio-Planck sample. The dashed line represents a linear fitlog(variability Index) = m × log(CD) + k, where the con-stants are m = 0.618 and k = 2.39.0815, and BZQJ 2136+0041, and all cases show variabilityon timescales at least lower than one month (given thatthey are detected within isolated month bins), i.e., shorterthan the monthly time bin described by the variability in-dex parameter. Further studies are necessary to investigatethe dependencies of transient and fast flares with respectto a time-bin smaller than that of a month, which we used.
47 46 45 44
43 42
12 14
16 18 20 22 24 26
Fig. 23.
SED for BZQJ 1224+2122, with blue bump tem-plate assuming z = 0.434. This source has the largest γ -rayvariability index, and we fit the IC component during var-ious flaring states. In the high-energy band, the blue linerefers to the 1FGL detection (integrating 2008-2009 data),the red line refers to the 3FGL detection (2008-2012 data),and the dark-green line represents a short and relativelybright flare reported by the MAGIC team Aleksi´c et al.(2011) with reported variability within few hours.Another example of strong γ -ray variability isBZQJ 1224+2122 (4C +21.35), whose SED is shown inFig. 23. We fit both Syn and IC components with a third or-der polynomial, listing fit parameters and their correspond-ing CD values in table 4. We note that the high-energy peakflux changes by one order of magnitude in between 1FGL(dark blue) and 2FGL catalogs (red). This kind of longtimescale variability (several months) is well representedby the variability index. When checking the light curve
14. Arsioli and G. Polenta: A complete sample of LSP blazars fully described in γ -rays available online we see that this source had a relativelysteady γ -ray emission during the first year of observationsby Fermi-LAT (08/2008 to 08/2009, corresponding to the1FGL SED; dark blue points), while it later underwent astrong activity when the photon flux varied by more thanone order of magnitude in the 0.1-500 GeV band.Since a more active state appears just after the inte-gration time used for building the 1FGL catalog, we knowthat the flaring activity is now smoothed over two and fouryears of integration time. These integration times are usedfor the 2FGL and 3FGL catalogs represented by red andlight green points, respectively, in the high-energy SED.However, when integrating data only during the brighteststate, it is possible to observe E >
100 GeV flux variabilityof ≈ one order of magnitude within hourly timescale, as forthe flaring episode reported by Aleksi´c et al. (2011). Thisis a clear example of how the log(CD) parameter can varywidely, from − γ -ray emission) to +0.6(when integrating over steady+flaring states), and reachingup to +1.3 (at the peak-flaring) as reported in Table 4. Table 4.
Inverse Compton peak parameters in various flar-ing states for 5BZQJ 1224+2122. First to last line in tablecorrespond to blue, red, and green curves, respectively, fromFig. 23.
Epoch log( ν peakIC ) log( ν f ν ) log(CD)1FGL 21.5 − − − − Overall, these examples suggest that γ -ray flaring statesare likely to generate the largest CD values in the tail ofthe distribution (Fig. 20, top). During flare episodes, thesources could be moving from a nearly steady multicom-ponent SSC + EC regime, to a short-lived EC-dominatedregime, which produce large amplitude variability (up tothree orders of magnitude) owing to the extra beaming fac-tor ∝ δ − α Dermer (1995) that is present in EC scenario.We will investigate that in a forthcoming paper. ν Synpeak plane
Here we report on the relation between log(CD) andlog( ν Synpeak ) that has been argued in literature (Fossati et al.1998; Nalewajko & Gupta 2017; Potter & Cotter 2013) toshow a relatively strong correlation. In fact, for our samplethe Pearson’s correlation coefficient r between log(CD) andlog( ν Synpeak ) is very weak, r ∼ − ν peak space, suggesting that the correlation between log(CD) andlog( ν peak ) parameters should be considered with great carein order to evaluate its dependence on selection effects whenbuilding the study sample.It is also important to keep in mind that CD estimatesreported for blazars with log( ν peak ) > Light curve for BZQJ 1224+2122 (3FGLJ 1224.9+2122): http://fermi.gsfc.nasa.gov/ssc/data/access/lat/4yr_catalog/ap_lcs/lightcurve_3FGLJ1224.9p2122.png : : : : : : º Synpeak ) [Hz] ¡ L og ( C D ) -1 : : : : : : º Synpeak ) [Hz] ¡ L og ( C D ) : : : : : : º Synpeak ) [Hz] ¡ L og ( C D ) Fig. 24.
Log(CD) vs. log( ν Synpeak ) plane for the Radio-Planck sample, probing almost three orders of magni-tude in ν Synpeak space and showing very weak correlationwith a large scatter. The black dashed line is a linear fitlog(CD) = − ν Synpeak )+3.92 .mentioned in literature (Acero et al. 2015; Ackermann et al.2015b; Arsioli et al. 2015; Chang et al. 2017) that HSPblazars on average have a hard γ -ray spectral slope with (cid:104) Γ (cid:105) ranging from 1.8 to 2.0, and therefore the IC peak ismost of the times out of reach for the Fermi-LAT, given itssensitivity window. For the brightest cases in which the ICpeak was probed by VHE Cherenkov observatories, there isstill the uncertainty introduced by absorption of VHE pho-tons due to pair creation when scattering EBL photons.Indeed, only a few HSP sources have their IC peak probedby VHE observatories with observations triggered by X-rayand γ -ray flaring states. This alone introduces a strong bias,given that no VHE blind sky survey is available.
6. Conclusions
The Radio-Planck sample includes 104 bright radio-selectedsources (f > γ -rays. The noninclusion of a fair fraction of 5BZcat sourcesin published Fermi-LAT catalogs motivated our search fornew γ -ray detections using 7.5 years of data as available atthe time of writing. The main results of our work can besummarized as follows. Out of 104 sources, 83 have counter-parts from FGL catalogs (all TS > >
20; 3 are new detections with TS in between 10to 20; 3 are new associations with 3FGL sources (from im-proved positioning with high-energy TS maps); 1 is a newdetection from solving γ -ray source-confusion; and 3 aretransients that were detected during short flaring episodes.Five sources remain undetected in the γ -ray band,all of which are optically identified blazars included inthe 5BZcat. Two objects have relatively high redshift(5BZQJ 2139+1423 at z = 2.427 and 5BZQJ 2007+4029at z = 1.736). The remaining three sources are5BZQJ 0927+3902 (z = 0.695), 5BZQJ 2022+6136(z = 0.228), 5BZQJ 1927+7358 (z = 0.302).We conclude that most of sources currently called γ -rayquiet blazars are actually associated with relevant γ -ray sig-natures, becoming evident by means of a dedicated case-by-case study of 7.5 yr of Fermi-LAT observations. At most, γ -
15. Arsioli and G. Polenta: A complete sample of LSP blazars fully described in γ -rays ray quiet blazars might be a very small fraction of the LSPpopulation, suggesting there is no urgent need to introducea new blazar class. From the five nondetections reported,two are high redshift sources, where absorption may hin-der a γ -ray signature. Another case (5BZQJ 0927+3902)is associated with a bright 3FGL source that dominatesthe region, such that we could not probe for source con-fusion. Finally, 5BZQJ 2022+6136 and 5BZQJ 1927+7358only showed hints of flaring activity and are probably un-der Fermi-LAT sensitivity.All new detections reported in this work contribute tosolve a small fraction of the extragalactic γ -ray backgroundinto point-like sources. We note that the presence of tran-sient sources, which are only detectable during short flar-ing episodes, could represent a non-negligible fraction ofthe MeV to GeV background. This would be a possible ap-proach to consider in future studies. We discuss examplesof how to extract refined and relevant γ -ray information byconsidering a multifrequency approach when searching fornew sources, showing that Fermi-LAT database is a largeresource still to be explored in detail.We study the CD distribution, showing that a singleGaussian function fails to describe the cases with largelog(CD) values at the tail of the histogram. There is in-deed a number of high log(CD) sources that are in excesswith respect to a single Gaussian fitting. We evaluate theimpact of fast γ -ray variability on the CD parameter, con-sidering 5BZQJ 1224+2122 as an example. We point outthree cases in which large CD values are observed dur-ing fast flaring states (BZQJ 0010+1058, BZUJ 0241-0815,and BZQJ 2136+0041), such that CD values can be oneto two orders of magnitude larger compared to those ob-tained during the steady and relatively faint γ -ray emis-sion. As follows, the absent correlation between log(CD)versus log(Var. Index) in the scatter plot from Fig. 22 showsthat the γ -ray variability index may not be the best toolto evaluate the relation connecting γ -ray flaring states andlarge log(CD) sources. Finally, we also evaluate the puta-tive correlation between log(CD) and log( ν Synpeak ) parame-ters, finding relatively weak evidence for that. The similar-ity between Syn and IC ν f ν peak distributions and the tightpeak ratio log( ν ICpeak / ν Synpeak ) distribution points to a domi-nant mechanism (either SSC or EC) to account for the ICcomponent in bright LSP blazars, otherwise we would havefound a large spread in the parameter space we probed. Anextensive evaluation testing SSC and EC scenarios is ex-plored with great detail in a parallel work (Arsioli & Chang2018), which is based on the Radio-Planck sample.Also, we showed a few examples for which the power-lawfitting parameters estimated for faint γ -ray blazars, whichwere detected with TS between 10 to 25 under 3FGL setupintegrating over 4.0 years of Pass7 data, are later confirmedwhen integrating over a larger exposure time of 7.5 years.This vindicates the importance and usefulness of reportingfaint γ -ray signatures in association with blazar counter-parts. Acknowledgements.
During this work, BA was supported by theBrazilian Scientific Program Ciˆencias sem Fronteiras - Cnpq, andlater by S˜ao Paulo Research Foundation (FAPESP) with grant n.2017/00517-4. We would like to thank Prof. Paolo Giommi for hiscomments along the preparation of this work, Prof. Marcelo M. Guzzoand Prof. Orlando L. G. Peres for the full support which allowedthe author partnership with FAPESP. We thanks IcraNet and Prof.Carlo Bianco for the cooperation granting access to Joshua Computer Cluster (Rome-Italy) for Fermi-LAT data reduction. We thank theCCJDR Data Center at IFGW Unicamp (Campinas-Brazil) where wealso performed Fermi-LAT data reduction at their Feynman Cluster.We thank SSDC, Space Science Data Center from Agenzia SpazialeItaliana; University La Sapienza of Rome, Department of Physics;And State University of Campinas - Unicamp, IFGW Department ofPhysics for hosting the author. We make use of archival data and bib-liographic information obtained from the NASA-IPAC ExtragalacticDatabase (NED), data, and software facilities from the SSDC.
References
Abdo, A. A., Ackermann, M., Agudo, I., et al. 2010, ApJ, 716, 30Abdo, A. A., Ackermann, M., Ajello, M., et al. 2010, TheAstrophysical Journal Supplement Series, 188, 405Acero, F., Ackermann, M., Ajello, M., et al. 2015, ApJS, 218, 23Acero, F., Donato, D., Ojha, R., et al. 2013, ApJ, 779, 133Ackermann, M., Ajello, M., Albert, A., et al. 2013, ApJ, 771, 57Ackermann, M., Ajello, M., Albert, A., et al. 2015a, ApJ, 799, 86Ackermann, M., Ajello, M., Albert, A., et al. 2016a, Physical ReviewLetters, 116, 151105Ackermann, M., Ajello, M., Albert, A., et al. 2012, Phys. Rev. D, 85,083007Ackermann, M., Ajello, M., Allafort, A., Antolini, E., & Atwood, e. a.2011, ApJ, 743, 171Ackermann, M., Ajello, M., Atwood, W. B., & Baldini, e. a. 2015b,ApJ, 810, 14Ackermann, M., Ajello, M., Atwood, W. B., et al. 2016b, ApJS, 222,5Ajello, M., Atwood, W. B., Baldini, L., et al. 2017, ApJS, 232, 18Aleksi´c, J., Antonelli, L. A., Antoranz, P., et al. 2011, ApJ, 730, L8Ando, S., Benoit-L´evy, A., & Komatsu, E. 2014, Phys. Rev. D, 90,023514Angelakis, E., Hovatta, T., Blinov, D., et al. 2016, MNRAS, 463, 3365Arsioli, B. & Chang, Y.-L. 2017, A&A, 598, A134Arsioli, B. & Chang, Y.-L. 2018,
Accepted for publication, April 92018, A&A, Ref. number: AA/2018/33005Arsioli, B., Fraga, B., Giommi, P., Padovani, P., & Marrese, P. M.2015, A&A, 579, A34Atwood, W., Albert, A., Baldini, L., et al. 2013, ArXiv e-printsAtwood, W. B., Abdo, A. A., Ackermann, M., et al. 2009, ApJ, 697,1071Blinov, D., Pavlidou, V., Papadakis, I., et al. 2015, MNRAS, 453,1669Chang, Y.-L., Arsioli, B., Giommi, P., & Padovani, P. 2017, A&A,598, A17Condon, J. J., Cotton, W. D., Greisen, E. W., et al. 1998, AJ, 115,1693Cuoco, A., Komatsu, E., & Siegal-Gaskins, J. M. 2012, Phys. Rev. D,86, 063004De Angelis, A., Tatischeff, V., Grenier, I. A., et al. 2017, ArXive-printsDermer, C. D. 1995, ApJ, 446, L63Di Mauro, M. & Donato, F. 2015, Phys. Rev. D, 91, 123001Di Mauro, M., Donato, F., Lamanna, G., Sanchez, D. A., & Serpico,P. D. 2014, ApJ, 786, 129Finke, J. D. 2013, The Astrophysical Journal, 763, 134Fornasa, M. & S´anchez-Conde, M. A. 2015, Phys. Rep., 598, 1Fossati, G., Maraschi, L., Celotti, A., Comastri, A., & Ghisellini,G. 1998, MNRAS, 299, 433Franceschini, A., Rodighiero, G., & Vaccari, M. 2008, A&A, 487,837Giommi, P., Padovani, P., Polenta, G., et al. 2012a, MNRAS, 420,2899Giommi, P., Polenta, G., L¨ahteenm¨aki, A., et al. 2012b, A&A, 541,A160Gregory, P. C., Scott, W. K., Douglas, K., & Condon, J. J. 1996,ApJS, 103, 427Grupe, D., Komossa, S., Leighly, K. M., & Page, K. L. 2010, ApJS,187, 64Inoue, Y. 2014, Fifth Fermi Symposium Proceedings, Nagoya, Japan.Lister, M. L., Aller, M. F., Aller, H. D., et al. 2015, ApJ, 810, L9Lister, M. L., Homan, D. C., Kadler, M., et al. 2009, ApJ, 696, L22MacDonald, N. R., Jorstad, S. G., & Marscher, A. P. 2017, ApJ,850, 87Massaro, E., Maselli, A., Leto, C., et al. 2015, Astrophysics andSpace Science, 357
16. Arsioli and G. Polenta: A complete sample of LSP blazars fully described in γ -rays Mattox, J. R., Bertsch, D. L., Chiang, J., et al. 1996, ApJ, 461, 396Meyer, E. T., Fossati, G., Georganopoulos, M., & Lister, M. L. 2012,The Astrophysical Journal Letters, 752, L4Nalewajko, K. & Gupta, M. 2017, A&A, 606, A44Padovani, P., Alexander, D. M., Assef, R. J., et al. 2017, A&A Rev.,25, 2Padovani, P. & Giommi, P. 1995, ApJ, 444, 567Paliya, V. S., Marcotulli, L., Ajello, M., et al. 2017, ApJ, 851, 33Planck Collaboration, Aatrokoski, J., Ade, P. A. R., et al. 2011,A&A, 536, A15Potter, W. J. 2018, MNRAS, 473, 4107Potter, W. J. & Cotter, G. 2013, Monthly Notices of the RoyalAstronomical Society, 431, 1840Prokhorov, D. A. & Churazov, E. M. 2014, A&A, 567, A93Tatischeff, V., Tavani, M., von Ballmoos, P., et al. 2016, inProc. SPIE, Vol. 9905, Space Telescopes and Instrumentation2016: Ultraviolet to Gamma Ray, 99052NVanden Berk, D. E., Richards, G. T., Bauer, A., et al. 2001, AJ,122, 549
17. Arsioli and G. Polenta: A complete sample of LSP blazars fully described in γ -rays Table 5.
Here we lits all 104 sources used for our studies. Column 5BZcat shows the blazar name according to Massaro et al.(2015), where BZQ stands for Flat Spectrum Radio Quasars, BZB for BL Lacs and BZU for still undefined-class blazars. Followingwe list coordinates R.A. and Dec. (J2000) for each source. Columns NVSS shows the radio counterpart according to the NVSScatalog (Condon et al. 1998). Column z corresponds to the redshift as reported in the 5BZcat Massaro et al. (2015), and fromthe NASA/IPAC Extragalactic Database (NED). Columns log( ν Synpeak ) and log( ν Synpeak ) are the fitting parameters referring to thepeak frequency from Syn and IC components measured in Hz. Columns log( νf Synν ) and log( νf ICν ) correspond to the Syn and ICpeak-power measured in erg/cm /s. All fitting parameters are given as a measure of the mean SED, considering all available data. ν ) log( ν f ν ) log( ν IC ) log( ν f ν − IC )5BZBJ0050-0929 12.67167 -9.48500 005041-092906 0 14.6 -11.0 22.7 -11.15BZQJ1512-0905 228.21056 -9.09995 151250-090600 0.360 13.1 -10.9 22.2 -9.915BZQJ2229-0832 337.41705 -8.54847 222940-083254 1.560 13.1 -11.2 21.8 -10.55BZQJ0607-0834 91.99875 -8.58056 060759-083450 0.870 12.1 -11.5 21.4 -11.05BZUJ0241-0815 40.27000 -8.25576 024104-081521 0.005 13.5 -10.1 20.9(?) -10.65BZQJ0808-0751 122.06473 -7.85275 080815-075109 1.837 13.0 -11.1 22.9 -10.75BZQJ0006-0623 1.55791 -6.39333 000613-062335 0.347 13.0 -11.1 20.3 -11.95BZQJ1256-0547 194.04652 -5.78931 125611-054720 0.536 12.8 -10.0 22.7 -10.05BZQJ2225-0457 336.44693 -4.95039 222547-045701 1.404 13.0 -10.8 21.7 -10.95BZQJ2218-0335 334.71683 -3.59358 221852-033537 0.901 12.3 -11.3 21.0 -11.75BZQJ1743-0350 265.99527 -3.83461 174358-035004 1.057 12.6 -11.3 21.0 -11.25BZBJ2134-0153 323.54294 -1.88812 213410-015317 1.283 12.8 -11.3 21.8 -11.55BZQJ0501-0159 75.30338 -1.98729 050112-015912 2.291 13.0 -11.4 21.4 -11.15BZQJ0339-0146 54.87891 -1.77667 033930-014635 0.805 12.6 -11.3 22.0 -11.15BZQJ0423-0120 65.81583 -1.34253 042315-012032 0.916 13.0 -10.6 22.1 -10.75BZUJ0725-0054 111.46125 -0.91556 072550-005458 0.128 13.5 -11.0 20.5 -11.25BZQJ0125-0005 21.37017 -0.09889 012528-000556 1.077 12.8 -11.6 20.3 -11.65BZQJ2136+0041 324.16080 0.69839 213638+004154 1.941 11.7 -11.6 21.2 -11.25BZQJ0108+0135 17.16154 1.58342 010838+013458 2.099 12.9 -11.2 22.4 -10.65BZQJ0739+0137 114.82513 1.61794 073918+013704 0.189 13.9 -10.9 21.6 -10.85BZUJ1058+0133 164.62338 1.56633 105829+013358 0.890 13.1 -10.9 22.3 -10.85BZQJ1229+0203 187.27792 2.05222 122906+020305 0.158 13.4 -10.0 20.8 -9.545BZQJ1549+0237 237.37267 2.61700 154929+023700 0.414 12.9 -11.2 22.0 -11.15BZBJ0825+0309 126.45958 3.15667 082550+030924 0.506 13.1 -11.1 21.5(?) -11.65BZQJ1222+0413 185.59375 4.22083 122222+041317 0.966 12.7 -11.2 20.8 -10.75BZUJ0433+0521 68.29620 5.35433 043311+052115 0.033 13.8 -10.2 19.8 -10.25BZQJ2123+0535 320.93549 5.58947 212344+053522 1.941 12.6 -11.7 21.6 -11.85BZQJ1038+0512 159.69492 5.20808 103846+051229 0.473 12.0 -11.8 20.8 -12.15BZQJ1550+0527 237.64696 5.45290 155035+052710 1.422 13.0 -11.4 21.8 -11.45BZQJ2148+0657 327.02252 6.96055 214805+065739 0.999 12.5 -10.8 20.5 -11.05BZUJ1751+0939 267.88675 9.65019 175132+093901 0.322 13.1 -10.8 21.9 -10.85BZBJ0757+0956 119.27766 9.94303 075706+095634 0.266 13.7 -10.8 20.8 -11.15BZQJ0309+1029 47.26500 10.48778 030903+102916 0.863 12.8 -11.2 21.7 -11.25BZQJ1608+1029 242.19249 10.48550 160846+102908 1.226 12.8 -11.3 21.6 -11.05BZQJ1504+1029 226.10408 10.49417 150425+102938 1.839 12.8 -11.6 22.9 -10.05BZQJ0010+1058 2.62917 10.97472 001030+105827 0.089 14.5 -10.7 20.5 -10.85BZBJ0449+1121 72.28196 11.35778 044907+112128 2.153 12.9 -11.5 21.8 -10.85BZQJ2232+1143 338.15167 11.73055 223236+114350 1.037 12.4 -11.2 21.3 -10.55BZQJ0750+1231 117.71687 12.51801 075052+123104 0.889 12.6 -11.1 21.2 -11.25BZQJ0530+1331 82.73508 13.53200 053056+133155 2.070 12.2 -11.5 21.4 -10.75BZUJ1415+1320 213.99500 13.34000 141558+132024 0.247 12.8 -11.0 20.5 -11.25BZQJ2139+1423 324.75546 14.39333 213901+142336 2.427 12.2 -11.7 - -5BZUJ0204+1514 31.21004 15.23639 020450+151411 0.833 12.6 -11.6 21.0 -11.25BZQJ2253+1608 343.49042 16.14805 225357+160853 0.859 13.1 -10.0 22.2 -9.205BZBJ0238+1636 39.66167 16.61639 023838+163658 0.940 13.0 -10.9 22.5 -10.45BZQJ2203+1725 330.86203 17.43006 220326+172548 1.076 13.3 -10.9 22.9 -10.95BZBJ0738+1742 114.53083 17.70528 073807+174219 0.424 13.5 -10.6 23.8 -11.05BZQJ0510+1800 77.50988 18.01155 051002+180041 0.416 13.3 -11.3 21.3 -11.15BZBJ0854+2006 133.70332 20.10833 085448+200630 0.306 13.6 -10.4 21.8 -10.85BZQJ1224+2122 186.22713 21.37972 122454+212247 0.434 13.1 -10.7 23.0 -10.05BZQJ0152+2207 28.07525 22.11881 015218+220707 1.320 12.9 -11.5 21.8 -11.55BZQJ1327+2210 201.75359 22.18060 132700+221050 1.398 12.6 -11.7 21.7 -11.05BZQJ0830+2410 127.71700 24.18333 083052+241058 0.939 12.6 -11.1 21.8 -11.05BZQJ1043+2408 160.78749 24.14306 104309+240835 0.560 12.9 -11.5 21.7 -11.65BZQJ0956+2515 149.20784 25.25446 095649+251515 0.712 12.7 -11.4 21.9 -11.4
18. Arsioli and G. Polenta: A complete sample of LSP blazars fully described in γ -rays Table 5. continued. ν ) log( ν f ν ) log( ν IC ) log( ν f ν − IC )5BZQJ2236+2826 339.09363 28.48261 223622+282858 0.790 13.0 -11.2 22.5 -10.95BZQJ0237+2848 39.46838 28.80250 023752+284809 1.206 12.9 -11.2 22.0 -10.75BZQJ1159+2914 179.88263 29.24556 115931+291444 0.729 13.5 -11.1 22.6 -10.85BZQJ2203+3145 330.81210 31.76056 220314+314538 0.295 13.4 -10.9 20.9 -11.15BZQJ0336+3218 54.12542 32.30806 033630+321829 1.259 12.8 -11.3 20.2 -10.55BZUJ1310+3220 197.61940 32.34550 131028+322044 0.997 13.1 -10.8 22.2 -10.85BZQJ1613+3412 243.42084 34.21333 161341+341247 1.397 12.3 -11.5 21.7 -11.65BZQJ1635+3808 248.81454 38.13458 163515+380804 1.814 12.5 -11.1 21.7 -10.35BZQJ1130+3815 172.72198 38.25514 113053+381519 1.733 12.3 -11.7 22.6 -11.65BZQJ0927+3902 141.76254 39.03914 092703+390220 0.695 12.1 -11.2 - -5BZQJ0555+3948 88.87837 39.81366 055530+394848 2.365 12.0 -11.7 20.8 -10.95BZBJ1653+3945 253.46750 39.76000 165352+394536 0.033 17.8 -10.2 25.2 -10.65BZQJ1640+3946 250.12347 39.77898 164029+394646 1.660 12.9 -11.9 22.7 -10.85BZQJ1642+3948 250.74506 39.81028 164258+394837 0.593 13.0 -10.6 21.7 -10.75BZQJ2007+4029 301.93726 40.49683 200744+402948 1.736 12.2 -11.6 - -5BZQJ0948+4039 147.23059 40.66239 094855+403944 1.249 12.3 -11.8 20.8 -11.25BZUJ0319+4130 * 49.95042 41.51167 031948+413042 0.018 13.0(?) -10.4 23.3 -10.45BZBJ2202+4216 330.68042 42.27750 220243+421640 0.069 13.6 -10.0 21.3 -10.35BZUJ0909+4253 137.38959 42.89613 090933+425347 0.670 12.9 -11.5 20.9 -11.75BZQJ0646+4451 101.63346 44.85461 064632+445116 3.396 11.6 -11.8 21.8(?) -11.55BZQJ0920+4441 140.24359 44.69833 092058+444153 2.190 12.5 -11.2 22.3 -10.65BZQJ2354+4553 358.59033 45.88445 235421+455304 1.992 12.2 -11.9 21.4 -11.75BZQJ0136+4751 24.24412 47.85809 013658+475129 0.859 13.3 -10.9 22.2 -10.75BZUJ1829+4844 277.38251 48.74628 182931+484446 0.695 13.0 -11.3 20.7 -11.15BZQJ1153+4931 178.35196 49.51911 115324+493109 0.334 12.9 -10.9 21.2 -11.15BZQJ0808+4950 122.16529 49.84347 080839+495036 1.432 12.0 -12.2 20.6 -11.75BZQJ0359+5057 59.87396 50.96394 035929+505750 1.512 12.1 -10.7 21.3 -10.45BZQJ2038+5119 309.65433 51.32018 203837+511913 1.686 12.5 -11.5 21.4 -11.05BZQJ1955+5131 298.92810 51.53015 195542+513149 1.214 12.7 -11.7 20.7 -11.45BZQJ1740+5211 265.15408 52.19528 174036+521143 1.381 13.2 -11.3 21.3 -10.85BZBJ1419+5423 214.94417 54.38722 141946+542315 0.153 13.7 -10.6 21.9 -11.55BZBJ1824+5651 276.02948 56.85042 182407+565101 0.663 13.2 -11.3 22.1 -11.05BZUJ0102+5824 15.69067 58.40310 010245+582411 ?0.644 12.6 -11.1 21.8 -10.95BZQJ2022+6136 305.52783 61.61633 202206+613658 0.228 12.9 -11.2 - -5BZBJ0958+6533 149.69666 65.56499 095847+653354 0.367 13.4 -11.0 21.2 -11.35BZQJ1849+6705 282.31833 67.09403 184915+670540 0.657 13.1 -10.9 22.5 -10.65BZQJ0228+6721 37.20854 67.35084 022850+672101 0.523 12.8 -11.2 21.1 -11.45BZQJ1642+6856 250.53271 68.94437 164207+685638 0.751 12.5 -11.6 20.1(?) -12.35BZBJ1806+6949 271.71124 69.82445 180650+694928 0.046 14.0 -10.6 21.7 -11.05BZQJ0841+7053 130.35153 70.89506 084124+705341 2.218 12.4 -11.3 20.1 -10.25BZBJ0721+7120 110.47208 71.34333 072153+712036 0.0 13.9 -10.3 23.3 -10.55BZQJ0217+7349 34.37833 73.82555 021730+734932 2.367 12.2 -11.6 20.5 -10.45BZQJ1927+7358 291.95209 73.96711 192748+735802 0.302 13.1 -10.8 - -5BZBJ2005+7752 301.37961 77.87861 200531+775243 0.342 13.2 -11.2 21.5 -11.35BZBJ1800+7828 270.19034 78.46778 180045+782805 0.680 13.5 -10.7 21.9 -10.95BZQJ1153+8058 178.30208 80.97476 115312+805829 1.250 12.6 -12.0 21.1 -11.95BZQJ2356+8152 359.09497 81.88118 235622+815252 1.344 12.8 -11.8 21.1 -11.5M87 187.70592 12.39111 123049+122321 0.0042 13.0 -10.5 18.5(?) -10.43C111 64.58867 38.02661 041820+380148 0.0485 13.3 -10.6 20.1 -10.0)5BZQJ2236+2826 339.09363 28.48261 223622+282858 0.790 13.0 -11.2 22.5 -10.95BZQJ0237+2848 39.46838 28.80250 023752+284809 1.206 12.9 -11.2 22.0 -10.75BZQJ1159+2914 179.88263 29.24556 115931+291444 0.729 13.5 -11.1 22.6 -10.85BZQJ2203+3145 330.81210 31.76056 220314+314538 0.295 13.4 -10.9 20.9 -11.15BZQJ0336+3218 54.12542 32.30806 033630+321829 1.259 12.8 -11.3 20.2 -10.55BZUJ1310+3220 197.61940 32.34550 131028+322044 0.997 13.1 -10.8 22.2 -10.85BZQJ1613+3412 243.42084 34.21333 161341+341247 1.397 12.3 -11.5 21.7 -11.65BZQJ1635+3808 248.81454 38.13458 163515+380804 1.814 12.5 -11.1 21.7 -10.35BZQJ1130+3815 172.72198 38.25514 113053+381519 1.733 12.3 -11.7 22.6 -11.65BZQJ0927+3902 141.76254 39.03914 092703+390220 0.695 12.1 -11.2 - -5BZQJ0555+3948 88.87837 39.81366 055530+394848 2.365 12.0 -11.7 20.8 -10.95BZBJ1653+3945 253.46750 39.76000 165352+394536 0.033 17.8 -10.2 25.2 -10.65BZQJ1640+3946 250.12347 39.77898 164029+394646 1.660 12.9 -11.9 22.7 -10.85BZQJ1642+3948 250.74506 39.81028 164258+394837 0.593 13.0 -10.6 21.7 -10.75BZQJ2007+4029 301.93726 40.49683 200744+402948 1.736 12.2 -11.6 - -5BZQJ0948+4039 147.23059 40.66239 094855+403944 1.249 12.3 -11.8 20.8 -11.25BZUJ0319+4130 * 49.95042 41.51167 031948+413042 0.018 13.0(?) -10.4 23.3 -10.45BZBJ2202+4216 330.68042 42.27750 220243+421640 0.069 13.6 -10.0 21.3 -10.35BZUJ0909+4253 137.38959 42.89613 090933+425347 0.670 12.9 -11.5 20.9 -11.75BZQJ0646+4451 101.63346 44.85461 064632+445116 3.396 11.6 -11.8 21.8(?) -11.55BZQJ0920+4441 140.24359 44.69833 092058+444153 2.190 12.5 -11.2 22.3 -10.65BZQJ2354+4553 358.59033 45.88445 235421+455304 1.992 12.2 -11.9 21.4 -11.75BZQJ0136+4751 24.24412 47.85809 013658+475129 0.859 13.3 -10.9 22.2 -10.75BZUJ1829+4844 277.38251 48.74628 182931+484446 0.695 13.0 -11.3 20.7 -11.15BZQJ1153+4931 178.35196 49.51911 115324+493109 0.334 12.9 -10.9 21.2 -11.15BZQJ0808+4950 122.16529 49.84347 080839+495036 1.432 12.0 -12.2 20.6 -11.75BZQJ0359+5057 59.87396 50.96394 035929+505750 1.512 12.1 -10.7 21.3 -10.45BZQJ2038+5119 309.65433 51.32018 203837+511913 1.686 12.5 -11.5 21.4 -11.05BZQJ1955+5131 298.92810 51.53015 195542+513149 1.214 12.7 -11.7 20.7 -11.45BZQJ1740+5211 265.15408 52.19528 174036+521143 1.381 13.2 -11.3 21.3 -10.85BZBJ1419+5423 214.94417 54.38722 141946+542315 0.153 13.7 -10.6 21.9 -11.55BZBJ1824+5651 276.02948 56.85042 182407+565101 0.663 13.2 -11.3 22.1 -11.05BZUJ0102+5824 15.69067 58.40310 010245+582411 ?0.644 12.6 -11.1 21.8 -10.95BZQJ2022+6136 305.52783 61.61633 202206+613658 0.228 12.9 -11.2 - -5BZBJ0958+6533 149.69666 65.56499 095847+653354 0.367 13.4 -11.0 21.2 -11.35BZQJ1849+6705 282.31833 67.09403 184915+670540 0.657 13.1 -10.9 22.5 -10.65BZQJ0228+6721 37.20854 67.35084 022850+672101 0.523 12.8 -11.2 21.1 -11.45BZQJ1642+6856 250.53271 68.94437 164207+685638 0.751 12.5 -11.6 20.1(?) -12.35BZBJ1806+6949 271.71124 69.82445 180650+694928 0.046 14.0 -10.6 21.7 -11.05BZQJ0841+7053 130.35153 70.89506 084124+705341 2.218 12.4 -11.3 20.1 -10.25BZBJ0721+7120 110.47208 71.34333 072153+712036 0.0 13.9 -10.3 23.3 -10.55BZQJ0217+7349 34.37833 73.82555 021730+734932 2.367 12.2 -11.6 20.5 -10.45BZQJ1927+7358 291.95209 73.96711 192748+735802 0.302 13.1 -10.8 - -5BZBJ2005+7752 301.37961 77.87861 200531+775243 0.342 13.2 -11.2 21.5 -11.35BZBJ1800+7828 270.19034 78.46778 180045+782805 0.680 13.5 -10.7 21.9 -10.95BZQJ1153+8058 178.30208 80.97476 115312+805829 1.250 12.6 -12.0 21.1 -11.95BZQJ2356+8152 359.09497 81.88118 235622+815252 1.344 12.8 -11.8 21.1 -11.5M87 187.70592 12.39111 123049+122321 0.0042 13.0 -10.5 18.5(?) -10.43C111 64.58867 38.02661 041820+380148 0.0485 13.3 -10.6 20.1 -10.0